Add Hetzner Cloud DataSource

The Hetzner Cloud metadata service is an AWS-style service available
over HTTP via the link local address 169.254.169.254.

https://hetzner.com/cloud
https://docs.hetzner.cloud/

8 files changed, 246 insertions, 4 deletions

diff --git a/cloudinit/apport.py b/cloudinit/apport.py
index 221f341c..618b0160 100644
--- a/cloudinit/apport.py
+++ b/cloudinit/apport.py
@@ -14,9 +14,9 @@ except ImportError:
 
 KNOWN_CLOUD_NAMES = [
     'Amazon - Ec2', 'AliYun', 'AltCloud', 'Azure', 'Bigstep', 'CloudSigma',
-    'CloudStack', 'DigitalOcean', 'GCE - Google Compute Engine', 'MAAS',
-    'NoCloud', 'OpenNebula', 'OpenStack', 'OVF', 'Scaleway', 'SmartOS',
-    'VMware', 'Other']
+    'CloudStack', 'DigitalOcean', 'GCE - Google Compute Engine',
+    'Hetzner Cloud', 'MAAS', 'NoCloud', 'OpenNebula', 'OpenStack', 'OVF',
+    'Scaleway', 'SmartOS', 'VMware', 'Other']
 
 # Potentially clear text collected logs
 CLOUDINIT_LOG = '/var/log/cloud-init.log'
diff --git a/cloudinit/settings.py b/cloudinit/settings.py
index c120498f..5fe749d4 100644
--- a/cloudinit/settings.py
+++ b/cloudinit/settings.py
@@ -36,6 +36,7 @@ CFG_BUILTIN = {
         'SmartOS',
         'Bigstep',
         'Scaleway',
+        'Hetzner',
         # At the end to act as a 'catch' when none of the above work...
         'None',
     ],
diff --git a/cloudinit/sources/DataSourceHetzner.py b/cloudinit/sources/DataSourceHetzner.py
new file mode 100644
index 00000000..769fe131
--- /dev/null
+++ b/cloudinit/sources/DataSourceHetzner.py
@@ -0,0 +1,100 @@
+# Author: Jonas Keidel <jonas.keidel@hetzner.com>
+# Author: Markus Schade <markus.schade@hetzner.com>
+#
+# This file is part of cloud-init. See LICENSE file for license information.
+#
+"""Hetzner Cloud API Documentation.
+   https://docs.hetzner.cloud/"""
+
+from cloudinit import log as logging
+from cloudinit import net as cloudnet
+from cloudinit import sources
+from cloudinit import util
+
+import cloudinit.sources.helpers.hetzner as hc_helper
+
+LOG = logging.getLogger(__name__)
+
+BASE_URL_V1 = 'http://169.254.169.254/hetzner/v1'
+
+BUILTIN_DS_CONFIG = {
+    'metadata_url': BASE_URL_V1 + '/metadata',
+    'userdata_url': BASE_URL_V1 + '/userdata',
+}
+
+MD_RETRIES = 60
+MD_TIMEOUT = 2
+MD_WAIT_RETRY = 2
+
+
+class DataSourceHetzner(sources.DataSource):
+    def __init__(self, sys_cfg, distro, paths):
+        sources.DataSource.__init__(self, sys_cfg, distro, paths)
+        self.distro = distro
+        self.metadata = dict()
+        self.ds_cfg = util.mergemanydict([
+            util.get_cfg_by_path(sys_cfg, ["datasource", "Hetzner"], {}),
+            BUILTIN_DS_CONFIG])
+        self.metadata_address = self.ds_cfg['metadata_url']
+        self.userdata_address = self.ds_cfg['userdata_url']
+        self.retries = self.ds_cfg.get('retries', MD_RETRIES)
+        self.timeout = self.ds_cfg.get('timeout', MD_TIMEOUT)
+        self.wait_retry = self.ds_cfg.get('wait_retry', MD_WAIT_RETRY)
+        self._network_config = None
+        self.dsmode = sources.DSMODE_NETWORK
+
+    def get_data(self):
+        nic = cloudnet.find_fallback_nic()
+        with cloudnet.EphemeralIPv4Network(nic, "169.254.0.1", 16,
+                                           "169.254.255.255"):
+            md = hc_helper.read_metadata(
+                self.metadata_address, timeout=self.timeout,
+                sec_between=self.wait_retry, retries=self.retries)
+            ud = hc_helper.read_userdata(
+                self.userdata_address, timeout=self.timeout,
+                sec_between=self.wait_retry, retries=self.retries)
+
+        self.userdata_raw = ud
+        self.metadata_full = md
+
+        """hostname is name provided by user at launch.  The API enforces
+        it is a valid hostname, but it is not guaranteed to be resolvable
+        in dns or fully qualified."""
+        self.metadata['instance-id'] = md['instance-id']
+        self.metadata['local-hostname'] = md['hostname']
+        self.metadata['network-config'] = md.get('network-config', None)
+        self.metadata['public-keys'] = md.get('public-keys', None)
+        self.vendordata_raw = md.get("vendor_data", None)
+
+        return True
+
+    @property
+    def network_config(self):
+        """Configure the networking. This needs to be done each boot, since
+           the IP information may have changed due to snapshot and/or
+           migration.
+        """
+
+        if self._network_config:
+            return self._network_config
+
+        _net_config = self.metadata['network-config']
+        if not _net_config:
+            raise Exception("Unable to get meta-data from server....")
+
+        self._network_config = _net_config
+
+        return self._network_config
+
+
+# Used to match classes to dependencies
+datasources = [
+    (DataSourceHetzner, (sources.DEP_FILESYSTEM, )),
+]
+
+
+# Return a list of data sources that match this set of dependencies
+def get_datasource_list(depends):
+    return sources.list_from_depends(depends, datasources)
+
+# vi: ts=4 expandtab
diff --git a/cloudinit/sources/helpers/hetzner.py b/cloudinit/sources/helpers/hetzner.py
new file mode 100644
index 00000000..2554530d
--- /dev/null
+++ b/cloudinit/sources/helpers/hetzner.py
@@ -0,0 +1,26 @@
+# Author: Jonas Keidel <jonas.keidel@hetzner.com>
+# Author: Markus Schade <markus.schade@hetzner.com>
+#
+# This file is part of cloud-init. See LICENSE file for license information.
+
+from cloudinit import log as logging
+from cloudinit import url_helper
+from cloudinit import util
+
+LOG = logging.getLogger(__name__)
+
+
+def read_metadata(url, timeout=2, sec_between=2, retries=30):
+    response = url_helper.readurl(url, timeout=timeout,
+                                  sec_between=sec_between, retries=retries)
+    if not response.ok():
+        raise RuntimeError("unable to read metadata at %s" % url)
+    return util.load_yaml(response.contents.decode())
+
+
+def read_userdata(url, timeout=2, sec_between=2, retries=30):
+    response = url_helper.readurl(url, timeout=timeout,
+                                  sec_between=sec_between, retries=retries)
+    if not response.ok():
+        raise RuntimeError("unable to read userdata at %s" % url)
+    return response.contents
diff --git a/tests/unittests/test_datasource/test_common.py b/tests/unittests/test_datasource/test_common.py
index 80b9c650..6d2dc5b5 100644
--- a/tests/unittests/test_datasource/test_common.py
+++ b/tests/unittests/test_datasource/test_common.py
@@ -14,6 +14,7 @@ from cloudinit.sources import (
     DataSourceDigitalOcean as DigitalOcean,
     DataSourceEc2 as Ec2,
     DataSourceGCE as GCE,
+    DataSourceHetzner as Hetzner,
     DataSourceMAAS as MAAS,
     DataSourceNoCloud as NoCloud,
     DataSourceOpenNebula as OpenNebula,
@@ -31,6 +32,7 @@ DEFAULT_LOCAL = [
     CloudSigma.DataSourceCloudSigma,
     ConfigDrive.DataSourceConfigDrive,
     DigitalOcean.DataSourceDigitalOcean,
+    Hetzner.DataSourceHetzner,
     NoCloud.DataSourceNoCloud,
     OpenNebula.DataSourceOpenNebula,
     OVF.DataSourceOVF,
diff --git a/tests/unittests/test_datasource/test_hetzner.py b/tests/unittests/test_datasource/test_hetzner.py
new file mode 100644
index 00000000..f1d1525e
--- /dev/null
+++ b/tests/unittests/test_datasource/test_hetzner.py
@@ -0,0 +1,99 @@
+# Copyright (C) 2018 Jonas Keidel
+#
+# Author: Jonas Keidel <jonas.keidel@hetzner.com>
+#
+# This file is part of cloud-init. See LICENSE file for license information.
+
+from cloudinit.sources import DataSourceHetzner
+from cloudinit import util, settings, helpers
+
+from cloudinit.tests.helpers import mock, CiTestCase
+
+METADATA = util.load_yaml("""
+hostname: cloudinit-test
+instance-id: 123456
+local-ipv4: ''
+network-config:
+  config:
+  - mac_address: 96:00:00:08:19:da
+    name: eth0
+    subnets:
+    - dns_nameservers:
+      - 213.133.99.99
+      - 213.133.100.100
+      - 213.133.98.98
+      ipv4: true
+      type: dhcp
+    type: physical
+  - name: eth0:0
+    subnets:
+    - address: 2a01:4f8:beef:beef::1/64
+      gateway: fe80::1
+      ipv6: true
+      routes:
+      - gateway: fe80::1%eth0
+        netmask: 0
+        network: '::'
+      type: static
+    type: physical
+  version: 1
+network-sysconfig: "DEVICE='eth0'\nTYPE=Ethernet\nBOOTPROTO=dhcp\n\
+  ONBOOT='yes'\nHWADDR=96:00:00:08:19:da\n\
+  IPV6INIT=yes\nIPV6ADDR=2a01:4f8:beef:beef::1/64\n\
+  IPV6_DEFAULTGW=fe80::1%eth0\nIPV6_AUTOCONF=no\n\
+  DNS1=213.133.99.99\nDNS2=213.133.100.100\n"
+public-ipv4: 192.168.0.1
+public-keys:
+- ssh-ed25519 \
+  AAAAC3Nzac1lZdI1NTE5AaaAIaFrcac0yVITsmRrmueq6MD0qYNKlEvW8O1Ib4nkhmWh \
+  test-key@workstation
+vendor_data: "test"
+""")
+
+USERDATA = b"""#cloud-config
+runcmd:
+- [touch, /root/cloud-init-worked ]
+"""
+
+
+class TestDataSourceHetzner(CiTestCase):
+    """
+    Test reading the meta-data
+    """
+    def setUp(self):
+        super(TestDataSourceHetzner, self).setUp()
+        self.tmp = self.tmp_dir()
+
+    def get_ds(self):
+        ds = DataSourceHetzner.DataSourceHetzner(
+            settings.CFG_BUILTIN, None, helpers.Paths({'run_dir': self.tmp}))
+        return ds
+
+    @mock.patch('cloudinit.net.EphemeralIPv4Network')
+    @mock.patch('cloudinit.net.find_fallback_nic')
+    @mock.patch('cloudinit.sources.helpers.hetzner.read_metadata')
+    @mock.patch('cloudinit.sources.helpers.hetzner.read_userdata')
+    def test_read_data(self, m_usermd, m_readmd, m_fallback_nic, m_net):
+        m_readmd.return_value = METADATA.copy()
+        m_usermd.return_value = USERDATA
+        m_fallback_nic.return_value = 'eth0'
+
+        ds = self.get_ds()
+        ret = ds.get_data()
+        self.assertTrue(ret)
+
+        m_net.assert_called_once_with(
+            'eth0', '169.254.0.1',
+            16, '169.254.255.255'
+        )
+
+        self.assertTrue(m_readmd.called)
+
+        self.assertEqual(METADATA.get('hostname'), ds.get_hostname())
+
+        self.assertEqual(METADATA.get('public-keys'),
+                         ds.get_public_ssh_keys())
+
+        self.assertIsInstance(ds.get_public_ssh_keys(), list)
+        self.assertEqual(ds.get_userdata_raw(), USERDATA)
+        self.assertEqual(ds.get_vendordata_raw(), METADATA.get('vendor_data'))
diff --git a/tests/unittests/test_ds_identify.py b/tests/unittests/test_ds_identify.py
index 9be3f964..9c5628e7 100644
--- a/tests/unittests/test_ds_identify.py
+++ b/tests/unittests/test_ds_identify.py
@@ -60,6 +60,7 @@ P_CHASSIS_ASSET_TAG = "sys/class/dmi/id/chassis_asset_tag"
 P_PRODUCT_NAME = "sys/class/dmi/id/product_name"
 P_PRODUCT_SERIAL = "sys/class/dmi/id/product_serial"
 P_PRODUCT_UUID = "sys/class/dmi/id/product_uuid"
+P_SYS_VENDOR = "sys/class/dmi/id/sys_vendor"
 P_SEED_DIR = "var/lib/cloud/seed"
 P_DSID_CFG = "etc/cloud/ds-identify.cfg"
 
@@ -379,6 +380,10 @@ class TestDsIdentify(CiTestCase):
         """Nocloud seed directory ubuntu core writable"""
         self._test_ds_found('NoCloud-seed-ubuntu-core')
 
+    def test_hetzner_found(self):
+        """Hetzner cloud is identified in sys_vendor."""
+        self._test_ds_found('Hetzner')
+
 
 def blkid_out(disks=None):
     """Convert a list of disk dictionaries into blkid content."""
@@ -559,6 +564,10 @@ VALID_CFG = {
              },
         ],
     },
+    'Hetzner': {
+        'ds': 'Hetzner',
+        'files': {P_SYS_VENDOR: 'Hetzner\n'},
+    },
 }
 
 # vi: ts=4 expandtab
diff --git a/tools/ds-identify b/tools/ds-identify
index ec368d58..e3f93c90 100755
--- a/tools/ds-identify
+++ b/tools/ds-identify
@@ -114,7 +114,7 @@ DI_DSNAME=""
 # be searched if there is no setting found in config.
 DI_DSLIST_DEFAULT="MAAS ConfigDrive NoCloud AltCloud Azure Bigstep \
 CloudSigma CloudStack DigitalOcean AliYun Ec2 GCE OpenNebula OpenStack \
-OVF SmartOS Scaleway"
+OVF SmartOS Scaleway Hetzner"
 DI_DSLIST=""
 DI_MODE=""
 DI_ON_FOUND=""
@@ -979,6 +979,11 @@ dscheck_Scaleway() {
     return ${DS_NOT_FOUND}
 }
 
+dscheck_Hetzner() {
+    dmi_sys_vendor_is Hetzner && return ${DS_FOUND}
+    return ${DS_NOT_FOUND}
+}
+
 collect_info() {
     read_virt
     read_pid1_product_name